linux-kernel - Re: [PATCH v2] trace: Add migrate-disabled counter to tracing output.

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20210906081343.klxtb652s5aphg4z@linutronix.de>
Date:   Mon, 6 Sep 2021 10:13:43 +0200
From:   Sebastian Andrzej Siewior <bigeasy@...utronix.de>
To:     Steven Rostedt <rostedt@...dmis.org>
Cc:     linux-kernel@...r.kernel.org, Ingo Molnar <mingo@...hat.com>,
        Thomas Gleixner <tglx@...utronix.de>,
        Peter Zijlstra <peterz@...radead.org>
Subject: Re: [PATCH v2] trace: Add migrate-disabled counter to tracing output.

On 2021-09-03 19:42:16 [-0400], Steven Rostedt wrote:
Hi Steven,

> BTW,
> 
> When doing a v2, always create a new thread. Never send it as a reply to
> the previous patch. The reason I missed this is because replies to previous
> patches do not end up in my internal patchwork. And I only look at that for
> patches. Not my INBOX.

oki.

> 
> > --- a/kernel/trace/trace_events.c
> > +++ b/kernel/trace/trace_events.c
> > @@ -181,7 +181,8 @@ static int trace_define_common_fields(void)
> >  
> >  	__common_field(unsigned short, type);
> >  	__common_field(unsigned char, flags);
> > -	__common_field(unsigned char, preempt_count);
> > +	/* XXX */
> > +	__common_field(unsigned char, preempt_mg_count);
> >  	__common_field(int, pid);
> >  
> >  	return ret;
> 
> I'm going to have to nuke this hunk of the patch, and update all the other
> locations that have preempt_mg_count in it. Because I just tested it, and
> this breaks user space.

I left that with that XXX on purpose so you can look and comment and say
how to fix it. The problem I had with preempt_count was that only the
lower nibble contains the preemption counter. So I thought you could
easily came up with something how this can be split or taught to only
expose the lower nibble. I didn't after a few attempts.
My understanding is that this `preempt_count' is also used in the filter
tracefs file so something like "preempt_count > 1" > filter would also
match for migrate_count = 2, preempt_count = 1. You seem to be happy to
filter in user space.

Sebastian